Research map / Video / Video segmentation and tracking
Robust Anomaly Detection in Videos Using Multilevel Representations
Hung Vu, Tu Dinh Nguyen, Trung Le, Wei Luo, Dinh Phung
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add top-layer DAE abstract detector to low-level detector (ablation) | MLAD0 · Avenue pixel-level | AUC: 47.07 -> 52.82 | Improved |
| add top-layer DAE abstract detector to low-level detector (ablation) | MLAD0 · UCSD Ped 2 pixel-level | AUC: 92.96 -> 94.45 | Improved |
| add top-layer DAE abstract detector to low-level detector (ablation) | MLAD0 · UCSD Ped 1 pixel-level | AUC: 66.07 -> 66.60 | No clear effect |
| add top-layer DAE abstract detector, relabeled ground truth (ablation) | MLAD0 · UCSD Ped 1* pixel-level | AUC: 64.41 -> 66.95 | Improved |
| add AlexNet Conv5 features instead of scene-trained DAE (ablation) | MLAD0 · Avenue pixel-level | AUC: 47.07 -> 40.60 | Hurt |
| add AlexNet Conv5 features (ablation) | MLAD0 · UCSD Ped 1 pixel-level | AUC: 66.07 -> 63.48 | Hurt |
| combine all or >=3 abstract-layer detectors instead of one (ablation) | MLAD0+3 · UCSD/Avenue | AUC | No clear effect |
| use a single abstract-level detector alone (ablation) | MLAD · UCSD Ped 1/2 | AUC | Hurt |
| use larger DAE network (ablation) | MLAD0+3 · UCSD Ped 2 | AUC: 97.52 -> 99.21 | Improved |
| MLAD vs best prior methods | MLAD0+3(A) · UCSD Ped 2 frame-level | AUC: 95.50 -> 97.52 | Improved |
Closest papers by meaning
- Learning from Noisy Supervision: A Denoising-Debiasing Framework for Weakly Supervised Video Anomaly Detection (CVPR 2026 · similarity 0.88)
- FrameShield: Adversarially Robust Video Anomaly Detection (NeurIPS 2025 · similarity 0.87)
- MULDE: Multiscale Log-Density Estimation via Denoising Score Matching for Video Anomaly Detection (CVPR 2024 · similarity 0.86)
- Holmes-VAU: Towards Long-term Video Anomaly Understanding at Any Granularity (CVPR 2025 · similarity 0.86)
- Open-Vocabulary Video Anomaly Detection (CVPR 2024 · similarity 0.86)
- ADPretrain: Advancing Industrial Anomaly Detection via Anomaly Representation Pretraining (NeurIPS 2025 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
